I’ve written an implementation of VOSIM (VOcal SIMulation),
a historically interesting synthesis technique that I long wanted to try out.
(See eg Dodge, Microsound 2nd ed, ch 8, or google)
Code, docs, example CSD are in http://abc.se/~re/Stuff/vosim.zip
Caveat: The code was developed and tested with Csound 4.x code, then converted
and just compiled with Csound5 sources.
So not fully tested – high risk of some errors there.
Just drop a note to the list if there’s a problem, it ought to be fixable…
VOSIM is a kind of pulse oscillator with a characteristic formant sound.
The sound can be tweaked some ways, but in all it’s not fantastic.
There are some limitations, eg formant bandwidth is not fully under user control.
Also there are audible steps in the spectral profile when parameters are changed.
OTOH it has a distinct early-computer-music period kinda sound!
Certainly worth a few trial runs but don’t expect to pass out with joy…
Disclaimer to above opinion: It would be interesting to try out some actual parameter settings
that the original authors developed for vocal simulation – I may not have found the sweet spots,
or the right way to set up groups of vosim instances.)